CircRNAs as biomarkers of cancer: a meta-analysis

BACKGROUND: The expression of circular RNA (circRNA) may affect tumor progression. However, there have been no systemic meta-analysis for cancer diagnosis by using circRNAs in clinical till now. Herein, we aimed to collect and examined all the evidence on the potential role of circRNA as novel biomarker in human cancers. METHODS: A comprehensive search strategy was used to search relevant literatures in the databases of PubMed, Embase, and the Web of Science from 2015 to August 2017. The correlation between circRNA expression and the diagnostic accuracy of tumor markers was analyzed. The methodological quality of each study was assessed by quality assessment for the diagnostic accuracies of the eligible studies (QUADAS-2). Statistical analysis was performed by applying the STATA (version 12.0) software. RESULTS: The present meta-analysis included 1752 patients with circRNA expression data of tumor and paired adjacent non-tumorous tissues from 17 publications (19 studies). The pooled sensitivity, specificity, positive likelihood ratios (PLR), negative likelihood ratios (NLR), and diagnostic OR (DOR) with their 95% confidential intervals (95%CIs), and AUC values were 0.72 (0.67–0.76), 0.74 (0.69–0.78), 2.80 (2.40–3.10), 0.38 (0.33–0.44), 7.00 (6.00–9.00), and 0.79, respectively. Subgroup analyses showed that the expression of circRNA in tissues of hepatocellular carcinoma (HCC) group was more prone to be detected than other tumor types, with a high values of the specificity, DOR, and AUC. CONCLUSIONS: circRNAs might be suitable as diagnostic biomarkers for tumors, especially in HCC diagnosis. Further prospective studies on the diagnostic value of circRNAs from the different tumors are needed in the future. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s12885-018-4213-0) contains supplementary material, which is available to authorized users.